381
J. Agric. Food Chem. 1994, 42, 387-387
Absorption, Tissue Distribution, Metabolism, and Excretion of Moxidectin in Cattle Jack Zulalian,' Steven J. Stout, Adrian R. d a c u n h a , Temistocles Garces, and Phillip Miller Metabolism Laboratories, Agricultural Research Division, American Cyanamid Company, P.O. Box 400, Princeton, New Jersey 08543-0400
The absorption, tissue distribution, metabolism, and excretion of moxidectin, a new endectocide for the control of internal and external parasites in cattle and sheep, was studied in cattle. Following a single subcutaneous dose of 14C-and 2H-labeled moxidectin of 0.2 mg/kg of body weight, highest 14C residues were present in abdominal fat (898,636, and 275 ppb) and back fat (495,424, and 186 ppb) a t 7, 14, and 28 days posttreatment, respectively. Lower residues were detected in liver (109, 77, and 31 ppb), kidney (42,38, and 13 ppb), and loin muscle (21,10, and 4 ppb), respectively. The administered radioactivity was excreted primarily in the feces, with only 3 % of the dose being eliminated in the urine. The HPLC/14C profiles of the residues extracted from the tissues, fat, and feces were qualitatively similar and showed moxidectin was the major component of the residue. Only two metabolites were present that were more than 5 % (2 ppb) of the total liver residues after 28 days. These were identified as the C-29/30 and the C-14 monohydroxymethyl metabolites by LC/MS and LC/MS/MS analysis of the metabolites isolated from the feces. Proton NMR analysis of the authentic compounds prepared in-uitro from cattle liver microsomal incubation and rat liver homogenate incubation with 14C-labeled moxidectin confirmed the mass spectral results. By LC/MS and LC/MS/MS, several other mono- and dihydroxylated and 0-demethylated metabolites were also identified.
Moxidectin is a semisynthetic derivative of nemadectin (Asato and France, 1990), a macrocyclic lactone produced by fermentation in a culture of Streptomyces cyanogriseus. It is active a t extremely low dosages against a wide variety of nematode and arthropod parasites and is currently being marketed in various countries as an injectable and a pouron product for beef cattle, as an oral drench and injectable product for sheep, and in tablet form for dogs. The chemical structure for moxidectin (Figure 1)is related to that of the milbemycins (Mishima et ai., 1983; Takiguchi et al., 1980) and avermectins (Albers-Schonberg et al., 1981), which have a novel mode of action against a broad spectrum of nematode and arthropod parasites of animals (Putter et al., 1981). The common features of these compounds are a fused cyclohexene-tetrahydrofuran ring system, a bicyclic 6,6-membered spiroketal, and a cyclohexene ring fused to the 16-membered macrocyclic ring. Moxidectin is the 23-(0-methyloxime) derivative of nemadectin and structurally differs from ivermectin (Fisher and Mrozik, 1989) in that it has no sugar moiety a t the C-13 position and has the unsaturated side chain at the C-25 position. The exact mechanism of action of moxidectin has not been fully determined, but it would appear to be multifaceted. y-Aminobutyric acid (GABA) is an important compound involved in the transmission of nerve signals and has a central role in this mode of action. Laboratory studies have demonstrated that moxidectin acts as an agonist to GABA receptors and stimulates the binding of this neurotransmitter. Paralysis of the parasite and eventual death and expulsion result. Such activity can also interfere with parasite reproduction as demonstrated by markedly reduced oviposition in engorged female ticks on cattle and sheep. Alternatively, the glutamate-chloride channel may also function as a central role in the mode of action of related milbemycins. Arena et al. (1992) have shown avermectin-sensitive chloride currents in Xenopus oocytes injected with Caenorhabditis elegans RNA. 002 1-856 119411442-038 1$04.5010
Com DOUnd
Moxidectin Metabolite A or 01'
or
X
l
H
H
n
H
on
OH OH H
n n
n
H H H H
H H
OH OH
H H
H OH
H H H
OH H
H
n
H H
n
OH
OH H
H
H
n
H
OH
H OH H OH
Metabolite C
n Metabolite D and 01'
Metabolite E
&
&
n
&
OH
on
~
H H
OH H OH
n
Figure 1. Metabolic pathway for moxidectin in cattle.
The absorption, tissue distribution, metabolism, and excretion of moxidectin were studied in cattle to evaluate the toxic potential of the residual tissue concentration of moxidectin and its metabolites. Comparative metabolism studies were also conducted in sheep (Afzal et al., 1994) and the laboratory rat (Wu et al., 1993). Comparative in-vitro studies were conducted with cattle liver microsomes and rat liver homogenates (Zulalian et d.,1992), and pharmacokinetic studies were conducted in cattle, sheep, and the laboratory rat. The cattle study described herein was conducted with 14C-labeledand deuteriumlabeled moxidectin following a single subcutaneous in@ 1994 American Chemical Society
382 J. Agric. Food Chem., Vol. 42, No. 2, 1994
jection of 0.2 mglkg of body weight. NMR and mass spectrometric examinations were carried out on the metabolites isolated from the in-vitro metabolism studies for confirmation of structure and comparison to the metabolites isolated from the cattle study. MATERIALS AND METHODS '%-Labeled Moxidectin. Nemadectin, the precursor to moxidectin, was multiply-labeled with 14C by incorporation of a mixture of carboxyl-l4C-1abeledacetate, propionate, and isobutyrate (Z. Ahmed and M. W. Bullock,AmericanCyanamid,private communication, 1991) and chemically modified using a semisynthetic reaction sequence to produce 1%-labeled moxidectin with a radiopurity of 96 % by HPLC and TLC analysis,a chemical purity of 97% by HPLC analysis, and a specific activity of 34.1 pCi/mg. 2H-Labeled Moxidectin. Moxidectin labeled at the 5-(u position with deuterium was prepared via oxidation of the C-5 hydroxyl group to a ketone followed by reduction with sodium borodeuteride (Ahmed and Bullock, private communication, 1991). The chemical purity was 95% as determined by HPLC. Dose Formulation. A formulation of 1% (w/v) active ingredient was prepared from a mixtureof 1%-labeled moxidectin and 2H-labeledand unlabeled moxidectin. The concentration of moxidectin in the injectable solution was 9.33 mg/g (161.3 f 4.1 pCi/g) as determined by HPLC analysis and liquid scintillation counting. The specific activity of moxidectin was 16.6 pCi/mg. The 2H-labeled moxidectin was used to serve as a mass marker in the identification of the metabolites by mass spectrometry. A blank formulation was prepared for treatment of the controlsteer. The density of the injectablesolutionfor the subcutaneous dosing was 1.04 g/mL. Handling and Treatment of Steers. Hereford steers weighing between 212 and 243 kg were used for the metabolism study. After a period of adaptation and observation for any clinical signsof disease,the steers were placed in individual pens and the activity of each steer was restricted (as would be in the metabolism stall) by the use of a halter attached to a rope tied to the stanchion. Animals were fed a roughage mixture (70% corn silage and 30% chopped alfalfa hay) ad libitum plus 1.0 kg of 20 % protein supplement per head per day. Water was supplied ad libitum. Three animals were randomly assigned to the treatment groupwith sacrifice times of 7,14,and 28 days postdose and one animal was assigned to the control with a sacrifice time of 6 days postdose. The steers were weighed on the day before and on the day they were placed in metabolism stalls, and the average weight of each steer on these two days was used for the determination of the drug dosage. Each steer in the treatment group received a single subcutaneous injection of the 1% injectable solution (4.29-5.02 g) equal to 0.2 mg of moxidectini kg of body weight. The control steer received 4.40 g of the blank formulation. The dose was administered anterior to the left shoulder using an 18-gauge, 1.5-in. needle. Collection of Samples. Total daily urine and feces outputs were collected and recorded for each steer starting 1 day prior to treatment and continuing until the time of scheduled sacrifice. Urine was collected from each steer using a canvas collection bag, patterned after one obtained from Colorado State University for use in cattle. The collection device was placed around the prepuce area of the steer and sutured with nylon to the adjacent abdominal skin. Tygon flexible plastic tubing was connected from the collectiondevice to a 5-galglassbottle for urine collection. A stainless steel collection pan lined with a sheet of plastic and placed on the floor at the back of the metabolism stall was used for collectionof the feces. The total daily collection was mixed thoroughly with a wooden paddle. The fecal mixture was then placed in a heavy-duty plastic bag, and the feces were further mixed by kneading the plastic bag over the feces. Approximately 2 kg of feces from each steer, randomly collected from different areas of the already well-mixed sample, was provided for radioanalysis. At sacrifice, blood, liver, kidney, fat (omental and back), skin with associated fat and muscle from an area surrounding the injection site (15cm wide and 2.5 cm deep), and loin muscle from the right side of the carcass were collectedfrom each steer. A number of other biological samples were also
Zulallan et al. collected as listed in Table I. Bile was collected directly from the gall bladder at sacrifice. Sample Preparation and Radioanalysis Procedures. All tissues, thoracic and abdominal viscera, and skin with associated fat and muscle were ground with dry ice and stored frozen, and the dry ice was allowed to dissipate. The radioactive residues were determined by homogenizing a 5.0-g aliquot of the ground sample with 5.0 g of water and combusting 1.0-g aliquota of the homogenate (equal to 0.5 g of tissue) in triplicate. Blood, bile, gastrointestinal contents, and feces were analyzed by combusting 0.5-g aliquots in triplicate. Urine was analyzed by counting 1.0mL aliquots directly in liquid scintillationcocktail. Radioactivity was measured in a Beckman LS 5801or Beckman LS 9800 liquid scintillation counter. Sample combustions were performed in a Model 306 Packard oxidizer. Extraction of W Residues. The extractions of the 14C residues from the tissues and the feces were performed in a homogenizer using 10mL of solvent/g of sample. The fat, edible tissues, and injection site samples were extracted (3X) with acetonitrile. The postextracted solids from the liver were extracted (2x)withmethanol/water (WlOv/v/). The feces were extracted (3X) with methanol/water (955 v/v). Urine samples (100-mL aliquota) were extracted with diethyl ether (4 X 150 mL) followed by methylene chloride (4 X 150 mL). The bile (40-80 mL) was extracted (4 X 100 mL) with diethyl ether. The acetonitrile extracts of the tissues, fat, and injection site samples were extracted (3x1 with hexane using a solvent ratio of 1:3 (v/v) hexane/acetonitrile for further cleanup. Aliquota of the extracts, the aqueous phase, and the postextracted solids were analyzed for radioactivity by liquid scintillation counting and combustion. The aqueous and organic fractions from the extractions of the tissues, feces, urine, and bile were concentrated preparatory to HPLC and TLC analysis. Bile and urine were also analyzed directly by HPLC. The methanol/water extracts of the feces were extracted with diethyl ether for isolation of the metabolites for mass spectral analysis. Samples of control feces, urine, and fat were fortified with 14C-labeledmoxidectin, and the fortified samples were processed as described for the treated samples. In-Vitro Experiments. Microsomes from steer liver were prepared according to the procedure of Miwa et al. (1982). The washed microsomal pellet was resuspended in 0.25 M sucrose and stored at -78 "C at a protein concentration of 32.1 mg/mL. Steer liver microsomes and a commercially availablepreparation of S9 liver homogenates from Aroclor 1254 induced rats were incubated with 14C-labeledmoxidectin to generate metabolites for isolation and identification by mass spectrometry and for comparison to the in-uiuo metabolites. Preparative-scale incubations were also conducted for the isolation and structural determination of the in-uitro metabolites by proton NMR analysis. For the preparative-scale incubation with steer liver microsomes, '*C-labeled moxidectin (1.0 mg, specificactivity 8.3 pCi/mg) in 3.0 mL of methanol was added to a vial containing nonlabeled moxidectin (14.3 mg) and Pluronic surfactant F-68 (BASFWyandotte Corp., Parsippany, NJ)and diluted to 4.0 mL with methanol. This solutionwas divided into four equal portions and placed in 250-mL flasks. To each flask were added 50 mL of 1 M potassium phosphate buffer (pH 7.4), steer liver microsomes equivalent to 120 mg of protein, 10 mL of 0.075 M glucose 6-phosphate (0.858 g/40 mL),5 mL of 5.7 nM 8-nicotinamide adenine dinucleotide phosphate (84.4 mg/20 mL), 100 units of type IX glucose-6-phosphatedehydrogenase, and 20 mL of distilled water. The samples were aerobicallyincubated with mechanical shaking at approximately 37 "C for 1h, after which time additional steer liver microsomes were added, equivalent to 120 mg of protein in 25 mL of 0.1 M potassium phosphate buffer (pH 7.4). For the experiment conducted with S9 rat liver homogenates, W-labeled moxidectin (28 mg, 0.31 pCi/mg) and the surfactant (0.56 g) were dissolved in methanol (6.0 mL). To four 125-mLflasks were added 1.0 mL of the methanol solution (4.7 mg of moxidectin) and 45 mL of a cofactor mix [prepared from 125 mL of 0.2 M sodium phosphate buffer (pH 7.4), 0.78 g of &nicotinamide adenine dinucleotide phosphate, 0.38 g of glucose 6-phosphate, 0.65 g of potassium chloride, and 0.42 g of magnesium chloridehexahydrate,diluted to 225 mL with water],
Tissue Distrlbutlon of Moxidectin in Cattle
5 mL of the rat liver homogenate, and 10mL of 0.15 M potassium chloride containing 0.1 g of 8-adeninine dinucleotide, reduced form. After incubation overnight at 37 "C, the samples from the respective experiments were extracted with diethyl ether. The ether extracts were concentrated to dryness. The residual radioactivity was taken up in methanol for radioassay and HPLC analysis in mobile phase systems I and 11. The extracts were chromatographed on CISreversed-phase TLC plates to separate moxidectin from the metabolites. The metabolites from the analytical-scale incubations were isolated as a mixture and analyzed by mass spectrometry. The metabolites from the preparative-scale incubations were separated by TLC on normalphase plates, purified by HPLC, and analyzed by proton NMR. TLC Procedures. For comparison of the radioactivityprofiles by TLC, the extracts of the feces and tissues were chromatographed on 20 X 20 cm precoated silica gel 60 plates (EM Laboratories Inc., Elmsford, NY) using the solvent system of methylene chloride/ethyl acetate/formic acid (3406080 v/v/v) and on 20 x 20 cm LKC18F reversed-phase plates (Whatman Chemical Separation, Inc., Clifton, NJ) using the solvent system of acetonitrile/water (180:20 v/v). Solvent migration was 15cm. An incremental multiple development TLC procedure was used to separate and isolate the metabolites using Whatman PLKCl8 plates developed with acetonitrile/water (360:40 v/v) and Whatman PK6F silica gel plates developed with ethyl acetate/hexane (250:lOOviv). Radioactive zones on the TLC plates were located by means of autoradiography on XAR double-coated medical X-ray film (Eastman Kodak, Rochester, NY). High-Performance Liquid Chromatography (HPLC). The moxidectin-derived residues in the urine, bile, and extracts of tissues, fat, injectionsite, and feces were routinely quantitated by HPLC on a 4.6 mm X 25 cm APEX CISreversed-phase column followed by liquid scintillation of the collected fractions. Mobile phase system I, a gradient of 4060 methanol/water to 9010 methanol/water over a period of 50 min followed by 9010 methanol/water for another 25 min, was used for these analyses. The flow rate was 1.0 mL/min, and the run time was 75 min. Fractions were collected at 0.5- or 1.0-min intervals in a fraction collectorand analyzed for radioactivity followingthe addition of 5 mL of liquid scintillation cocktail to each vial. The column eluate was monitored at 242 nm and the UV profile recorded. A 4.6 mm X 25 cm SUPELCOSIL Ce-DB reversed-phase columnand mobile phase system 11,a combined isocratic-gradient mobile phase of methanol/water, both containing 0.1 M ammonium acetate,were also used for the chromatographic comparison ofthe radioactivityprofiies of the in-uiuoandin-vitro metabolites. The HPLC run time was 75 min. For the first 20 min the mobile phase was 7030:O.l M ammonium acetate. This was followed by two gradients. The first gradient was 25 min to 75:25:0.1 M ammonium acetate, and the second was 15 min to 90100.1 M ammonium acetate. For the final 15 min, the mobile phase was 9O:lOO.l M ammonium acetate. The in-vitro metabolites were purified by HPLC on the CSDB reversed-phase column using mobile phase system 111,70:30 methanol/water for 20 min followed by a gradient over 25 min to 7525 methanol/water. Fractions were collected manually or with the aid of the fraction collector. The isolates were concentrated to dryness in a SpeedVac SC 100 (Savant Instruments, Inc., Farmingdale, NY) and submitted for proton NMR analysis. Isolation of Metabolites. The ether extract of the methanol/ water extract of the steer feces (day 2) was concentrated to dryness. The residue was taken up in 7.5 mL of methanol and applied to three Whatman PLKCIS reversed-phase silica gel plates. The plates were developed in four stages using an incremental multiple development with acetonitrile/water (360 40 v/v) wherein the solvent was allowed to migrate only a prescribed distance from the spotting origin at each development stage. The solvent migration distances were 3.75,7.5, 12.5, and 15 cm, respectively. Following autoradiography of the TLC plates, the silica gel scrapings of bands of similar Rf (9.0,8.5,8.0, and 5.5) from each plate were combined and extracted with acetone to recover the radioactivity. The reversed-phase TLC isolate Rf 9.0 was further resolved into three components (R, 4.2, 3.8, and 2.8) following chroma-
J. Agric. Food Chem., Vol. 42, No. 2, 1994 383
tography on Whatman PK6F silica gel plates developed with ethyl acetate/hexane (2501OOv/v) using the incremental multiple development procedure. The isolate Rf 2.8 from this normalphase TLC was the major component of this mixture. The reversed-phase TLC isolate Rf 8.0 was resolved into four components (Rf5.5, 5.0, 4.5, and 2.8) after normal-phase TLC. The isolate of Rf 5.0 was the major component of the mixture. The two remaining reversed-phase TLC isolates (Rf8.5 and 5.5) each appeared as one component, with Rf2.8and 7.2, respectively, after normal-phase TLC. The normal-phase TLC isolates were analyzed using HPLC mobile phase systems I and I1 for correlation of the isolated metabolites to the components in the radiometabolite mixture followed by mass spectrometry. The radioactivityfrom the analytical-scalein-vitroexperiments were resolved into two discrete bands of radioactivity (Rf3.5 and 6.0) followingchromatography on reversed-phase silicagel plates developed with a single pass of acetonitrile/water (180:20 v/v). The Rf 3.5 isolate was moxidectin. The Rf 6.0 isolate contained a mixture of metabolites and was analyzed by LC/mass spectrometry. The radioactivity from the preparative-scale in-vitro experiments was resolved into four discrete bands of radioactivity (Rf9.0,7.0,6.0, and 3.5) following chromatography on reversedphase silica gel plates. The three zones of radioactivity due to the metabolites (Rf 9.0, 7.0, and 6.0) were isolated. The two minor isolates (Rf 7.0 and 9.0) were examined by mass spectrometry without further purification. The major isolate (Rf6.0) was further resolved into nine discreteradioactivebands following chromatography on normal-phase silica gel plates developedwith ethylacetate/hexane(2501OOv/v)using the incrementalmultiple development procedure. The three major bands of radioactivity from the normal-phase TLC, located at Rf 5.5, 5.0, and 4.8, respectively,were isolated for HPLC cleanup and NMR analysis. The minor bands, located at Rf 7.0, 6.5, 5.9, 4.0, 3.0, and 2.0, respectively, were isolated for mass spectral analysis. Mass Spectrometry. Structural characterizations of the metabolites by thermospray liquid chromatography/mass spectrometry (LC/MS) and thermospray liquid chromatography/ tandem mass spectrometry (LC/MS/MS) were performed on a Finnigan-MATTSQ 70 triple-stage quadrupole system equipped with a Finnigan-MAT thermospray accessory. The detailed conditions for the LC/MS and LC/MS/MS are described elsewhere (Stout et al., 1994). Proton NMR. NMR spectra were obtained on a Bruker AM500 NMR spectrometer in deuteriated chloroform. Chemical shifts were referenced with respect to the chloroform resonance in the spectrum placed at 7.26 ppm. This implies that the shifts relative to TMS were -0 ppm. The instrument tuning and shimmingwere optimized to yield good-qualityspectra. For data acquisition a 30° flip pulse was used. A relaxation delay of 3 s was employed. Typical 1024 scans were coherently signal averaged using 32K complex data points both spin time and the frequency domains, respectively. RESULTS AND DISCUSSION
Distribution and Excretion of Radioactive Residues. The major edible tissueslorgans for the meatconsuming public are the liver, kidney, muscle, and fat. These are also the edible tissues/organs specifically monitored by the U.S. FDA on animal health drugs. Of the tissues listed in Table 1,following the subcutaneous dosing of steer with 14C-labeled moxidectin equal to 0.2 mg/kg of body weight, the highest residue levels, expressed as ppb equivalents of moxidectin, were detected in the omental fat and back fat with depletion half-lives of 12 and 14 days, respectively. Lower residues were detected in the liver, kidney, and muscle. Muscle contained the lowest residue among the edible tissues. The steady decline in the magnitude of the '4c residues at each sacrifice interval provides convincing evidence for the lack of accumulation of the residues in the edible tissues. The depletion half-lives for the residues in the other edible tissues were 9 days for muscle and 11 days for liver and kidney. The lowest residues were detected in the brain,
384
J. Agric. Food Chem., Vol. 42, No. 2, 1994
Zulallan et al.
Table 1. Total 1% Residues in Tissues, Blood, and Bile from Steers Dosed Subcutaneously with I%-Labeled and aH-LabeledMoxidectin, Expressed as Parts per Billion Equivalents of Moxidectin tissue abdominal fat back fat kidney liver muscle adrenals bile bladder blood brain carcass esophagus GI contents heart injection site intestines (large) intestines (small) lungs pancreas stomach compartments spleen thyroid thymus tongue
7 898 495 42 109 21 88 159 33 10 7 115 80 22 94 383 139 111 32 83 74 8 403 197 120
days posttreatment 14 28 636 275 424 186 38 13 31 77 10 4 120 29 82 42 61 21 7 3 3 c2 42 67 67 61 2 17 37 91 192 96 167 65 100 19 161 12 69 32 42 30 26 10 90 127 83 48 61 90
indicating that the migration of the moxidectin-derived residues in blood to the brain was minimal. For the subcutaneously treated steers, the major route of excretion of the administered radioactive dose was via the feces, 32.2% after 7 days, 41.3% after 14 days, and 58.1% after 28 days. The excretion of radioactivity in feces peaked within 3 days ( 5 4 % of the dose). Low levels of radioactivity (1-3% of the dose) were excreted daily thereafter. Only 3 % of the radioactivity was recovered in the urine 28 days postdose. The overall recoveries of the administered radioactivity were 70.7-76.9% after all of the collected repositories in the steer were analyzed. The residual body parts (head, feet, and tail) and bone knuckles were not analyzed. Extraction and Residue Profiles in Tissues. The radioactive residues in the edible tissues, fat, and injection site were essentially all extractable (89-99% 1, and very little remained in the postextracted solids after 28 days (11% ,
